Archaeological monitoring in the streets of : recent discoveries

by Michael Kimber

with a contribution by Julie Franklin and illustrations by Elin Evertsson and Anna Faras Pagowska

Headland Archaeology Ltd, 13 Jane St, Edinburgh EH6 5HE

Scottish Archaeological Internet Report [30], 2009 www.sair.org.uk Published by the Society of Antiquaries of , www.socantscot.org.uk with Historic Scotland, www.historic-scotland.gov.uk and the Council for British Archaeology, www.britarch.ac.uk Editor Helen Bleck Produced by Archetype Information Technology Ltd, www.archetype-it.com ISBN: 978 09039 0399 8 ISSN: 1473-3803

Requests for permission to reproduce material from a SAIR report should be sent to the Director of the Society of Antiquaries of Scotland, as well as to the author, illustrator, photographer or other copyright holder.

Copyright in any of the Scottish Archaeological Internet Reports series (text and illustrations) rests with the SAIR Consortium and the individual authors.

The consent does not extend to copying for general distribution, advertising or promotional purposes, the creation of new collective works or resale. Contents

1 Abstract...... 1

2 Introduction ...... 2

3 Historical background ...... 4 3.1 Roman ...... 4 3.2 The medieval period and later ...... 4 3.2.1 Morphology of the burgh...... 4

4 Summary of key results ...... 8 4.1 Interventions around known sites ...... 8 4.1.1 The vicus ...... 8 4.1.2 The Newbigging pottery...... 9 4.1.3 The mill lade ...... 9

5 Distribution of archaeological deposits ...... 12 5.1 Musselburgh: High Street, Millhill, Kerr’s Wynd, Mansfield Avenue, Kilwinning Place, Newbigging ...... 12 5.2 : North High Street, Market Street, Bridge Street...... 12

6 Artefacts and their distribution (Julie Franklin) ...... 14 6.1 Roman finds...... 14 6.2 Medieval and post-medieval finds...... 14

7 Discussion and interpretation ...... 15 7.1 The archaeology of Musselburgh...... 15 7.1.1 Distribution of deposits ...... 15 7.1.2 Medieval Musselburgh...... 15 7.1.3 Fisherrow...... 15 7.2 Pipeline transects in historic burghs ...... 15 7.2.1 Reasons for variation – population and development...... 16 7.2.2 Reasons for variation – geography...... 16

8 Conclusion ...... 17

9 References ...... 18

iii LIST OF ILLUSTRATIONS

1 Location of monitored interventions ...... 3 2 The development of Musselburgh from 1682–1816...... 5 3 Musselburgh in the 1850s (Source: Ordnance Survey 1854 sheets III & VII)...... 6 4 Section of Roman ditch on Inveresk Brae...... 8 5 Location of key sites and archaeological deposits ...... 10 6 Sherd of Samian ware recovered from T200/012...... 14

iv LIST OF TABLES

1 Distribution of archaeological deposits: Musselburgh ...... 13 2 Distribution of archaeological deposits: Fisherrow...... 13

 1 ABSTRACT

Archaeological monitoring of water mains renewal expected, they were fragmentary in comparison with in Musselburgh has provided new information on similar deposits from pipeline monitoring schemes the medieval and post-medieval development of the in Perth, and Crail. The data from burgh, as well as adding to known information on Musselburgh are in part less coherent due to the the vicus of the Roman fort, the Newbigging pottery kinds of work monitored within the burgh core, but and the town mill lade. Activity associated with the it is probable that they also reflect a lesser degree of Newbigging pottery seems to have extended further preservation of archaeological deposits beneath the to the west than the boundaries of the pottery road surface. This is partly due to modern develop- indicated on 19th-century maps of the town, while ment, and partly due to the geographic situation of Roman remains associated with the vicus survive the burgh, which does not appear to have encouraged in places beneath the road surface of Inveresk Brae. the formation of stratified deposits sealed by wind- However, while archaeological deposits related to blown sand, as in North Berwick, or the anaerobic the medieval burgh were located broadly where preservation conditions prevalent within Perth.

 2 INTRODUCTION

The renewal of water mains within Musselburgh by medieval burgh – was undertaken without an archae- Scottish Water between March 2003 and May 2004 ologist present and thus was largely unrecorded. provided an opportunity for archaeological monitor- The archaeological remains located in the course ing of extensive groundworks within and adjacent of the monitoring fall into two categories. The first is to the medieval core of the burgh, and adjacent to that of deposits or structures that can be related to the vicus of the Roman fort in Inveresk. Headland previously known sites. The new information about Archaeology Ltd was commissioned by Scottish these three sites – the vicus of Inveresk Roman Water’s contractor, Morgan Utilities, to monitor the Fort, the Newbigging pottery and the town mill lade works (Kimber 2004). – is considered first, along with the implications The mains replacement involved both open-cut of future development in their vicinity. The second trenching and smaller interventions in the roads and category of remains – archaeological structures and pavements of large parts of the burgh (illus 1). Guided deposits related to the growth and development of by the area of archaeological potential suggested the burgh – is discussed following this, and consid- by the Scottish Burgh Survey for Musselburgh ered in relation to the development of the burgh (Dennison & Coleman 1996), the monitoring covered from the medieval period. most streets within the town centre. In Inveresk, The final section of this article attempts to explain monitoring was extended outwith this area along the much more fragmentary evidence for medieval Inveresk Village Road and Carberry Road due to and later deposits recorded in Musselburgh, as the close proximity to the road of Scheduled areas compared with the far better preserved sequences including field systems, prehistoric settlement and recorded from similar projects in North Berwick Roman camps (SAMs 3610 & 3611). Archaeological (Dingwall 2004, Dingwall forthcoming), Crail (Lowe monitoring of the renewal works began following the 2001) and Perth (Glendinning 2002). The implica- discovery of a stone culvert on Carberry Road at the tions this comparison has for the preservation of instigation of the Council archaeologist, archaeological horizons within Musselburgh and for Biddy Simpson. Most of the open-cut trenching on the methodology of future fieldwork there will then Musselburgh High Street – one of the key areas of the be discussed.



BeechBeach Lane

Balcarres Eskside Rd Balcarres 334000 East Goose Green Pl RoRd 335000 673000 673000

Downie Pl Links Str James St Fishers WWynd

WWatt's Close New St Lochend Rd Nrth att's Harbour Rd est

ynd

North High St

Rd Sth Rd

Rd Sth Rd Eskside WWest East Port

South St St St South South Lochend Lochend North High St Linkfield RoRd MillMillhill Hi Kerrs WWyndynd Bridge St

Shorthope Hercus

Loan St Market St Mercat KillwinningKilwinning StSt

Newhailes Rd Rd

Rd New Bridge Campie

Campie High SSt Cross FisherFisherrow Row est KillwinningKilwinning Pl Musselburgh WWestest Holmes GardGdns Kilwinning Kilwinning Eskside WWest Terr T

ve Mansfield Ave err A Mansfield 335000 Mall Ave Mansfield 672500 Old Bridge Rd

A A ve venue

Olive Bank Station Fire Station Inveravon Rd Rd TerrT Newbigging err InvereskInvereskRd

Eskmills Est

ls g

Eskmil SAM 3267 Inveresk Brae

SAM 3285

SAM 1182 SAM 3612 334000 335000 672000 Inveresk VVillage Rd 672000 SAM 3293 SAM 3285 0.00.20.40.60.81.0 illage Ro

s

Double Dyke

Inveresk

Carberry RoRd SAM 36136111 SAM 36136111 River Esk 335000 671500

Open cut trenching N SAM 36113611 Pipe bursting/ sliplining only

Scheduled Ancient Monument

0 500 m

SAM data provided by the RCAHMS dated 061107 © Crown Copyright (RCAHMS) Reproduced using digital 1:1000 data supplied by Scottish Water/Morgan Utilities SAM 3610 with the permission of the Controller of HMSO. © Crown copyright. Headland Archaeology Ltd, licence no AL 100013329

Illus 1 Location of monitored interventions

 3 HISTORICAL BACKGROUND

Musselburgh is situated on a raised beach at the 3.2 The medieval period and later mouth of the River Esk. Historically, Musselburgh proper is that part of the town that grew up on Modern-day Musselburgh has origins at least as early the east bank of the Esk, while the settlement on as the 11th century ad, as a possession of Dunfermline the west bank is known as Fisherrow. The small Abbey. The elevated position of the church overlook- and affluent village of Inveresk occupies an area ing Musselburgh – by no means unique for a Scottish of higher ground overlooking Musselburgh from burgh (Mair 1988, 59–60) – may relate to the burgh’s the south, and originally grew up along the road original status. It is possible that a visible display of to the east of St Michael’s parish church. Between the Church’s relationship to the inhabitants was felt Musselburgh and Inveresk lies the area known as desirable by the ecclesiastical authorities. Newbigging, originally a separate hamlet, but now Indirect documentary evidence suggests that Mus- subsumed within Musselburgh. selburgh was granted burghal status in the later 12th century, but the first direct evidence of the special privileges associated with this status dates to the 3.1 Roman Inveresk early 14th century (Dennison & Coleman 1996, 19). As the Scottish state grew through the medieval The first major settlement around Musselburgh period and as its relationship with England waxed and was the Antonine-period Roman fort and vicus at waned, Musselburgh suffered from English military Inveresk. The development of the area around the campaigns and was burnt at least once. The area was parish church, which is in fact situated practically a target of raids during the ‘rough wooing’ and in 1547 in the centre of the fort (Leslie 2002), has thrown the was fought nearby. Among later up many examples of Roman remains over the past military impositions was the occupation of Mussel- several centuries. For example, it was reported burgh by Oliver Cromwell in 1650 as his army pressed in the Old Statistical Account (OSA) for Inveresk on Edinburgh following his victory at . that in the late 18th century, workmen in the The burgh itself, while developing textile, leather­ grounds of Inveresk House uncovered the floors of working and brewing industries from the 13th several Roman buildings forming part of the civil century onwards, seems not to have diversified settlement, including the remains of a bath-house, further, or gained a native class of professionals although at least some of these remains were later (Dennison & Coleman 1996, 36). This undoubt- destroyed in the course of the work that uncovered edly relates to the distorting effect Edinburgh was them (OSA 1791–99, 4). However, the remains of having upon the markets and social structure of a hypocausted building at the eastern edge of the its smaller neighbours – for example, by the 16th grounds of Inveresk House survive today, and were century Edinburgh, through its port of Leith, was re-excavated in the 1980s (Thomas 1988b), along already beginning to dominate the export trade with other parts of the civil settlement at Inveresk from the area, and its growing population was Gate, excavated in the 1970s (Thomas 1988a). making it increasingly attractive to both importers Further investigations in this area in the 1990s and exporters (Lynch 1988, 279). The proximity of produced finds that suggest occupation of the Musselburgh to Edinburgh and its markets brought civil settlement continued past the end of Roman benefits in the form of the settlement of the area military occupation of the area (Bishop 2002, 34). by the gentrified élite and the Edinburgh literati in Along with evidence for post-Roman activity in the the 18th and 19th centuries, and the construction field systems immediately south-east of Inveresk of new industries in the burgh, such as the Newbig- (Cook 2004), this offers clues that the area may ging pottery and Brunton’s Wire Works (Dennison & have continued as a focus for settlement up until Coleman 1996, 41). The great expansion of housing its emergence in historical records in the medieval in Musselburgh in the 20th century can in part be period. seen as the culmination of its success in becoming Information concerning the extent to which integrated with its dominant neighbour while Roman settlement extended west of the Esk is retaining much of its distinct identity. almost completely lacking. The OSA mentions a tradition that the remains of Roman buildings were frequently uncovered in Fisherrow in the 3.2.1 Morphology of the burgh vicinity of the harbour (OSA 1791–99, 5). Evidence for the occurrence of Roman remains in this area It has been suggested that the core of the medieval has yet to be confirmed by modern work. settlement was located near to the medieval or ‘Old

 Knox 1816 Roy ca 1750 Pinkie House Pinkie House Inveresk Pinkie House Musselburgh Inveresk Musselburgh Musselburgh Fisherrow River Esk Fisherrow River Esk River Esk Firth of Forth Harbour Harbour Firth of Forth Adair 1682 Laurie 1763 Pinkie House Musselburgh Pinkie House Illus 2 The development of Musselburgh from 1682–1816 The Illus 2 Musselburgh Inveresk Inveresk Fisherrow River Esk River Esk Fisherrow Firth of Forth Firth of Forth Harbour Harbour

 N

0 50 m

Illus 3 Musselburgh in the 1850s (source: Ordnance Survey 1854, sheets III & VII). Reproduced by permission of the Trustees of the National Library of Scotland.

Bridge’ over the Esk, and then expanded eastwards published (for example Hastie 2001) have generally along the line of the High Street throughout the produced negative results, in part reflecting the medieval period (Dennison & Coleman 1996, 52). modern development that has taken place in this Excavations at Brunton’s Wire Works, in the vicinity area. of the Old Bridge, recovered 13th- to 15th-century The first detailed map of Musselburgh, Fisherrow pottery from what have been described as ‘agricul- and Inveresk is Adair’s 1682 map of tural deposits’ (Ewart & Triscott 1993). Interventions (illus 2). In general, the core street plan in existence around the High Street since the Burgh Survey was then still forms the basis of the modern-day layout.

 In Musselburgh, the High Street is aligned east– Laurie’s map of 1763 shows the area in rather west, with Newbigging running south from the High more detail, confirming the existence by this time Street, and the road to the Old Bridge bypassing the of Hercus Loan, Lochend Road South and New southern edge of the High Street backlands. The Street in Fisherrow and Millhill in Musselburgh. focus of settlement at this time appears to have been One notable addition to the post-medieval street the intersection of the High Street and Newbigging, plan of Fisherrow is Bridge Street, first shown on where the mercat cross is located today. Mussel- Knox’s map of 1816. This street connects North burgh Links is marked on the map to the north of High Street in Fisherrow and High Street in Mus- the High Street. Settlement in Inveresk lies along selburgh, via the New Bridge which was built in the road running east from the parish church. In 1806–07 and runs through what were formerly the Fisherrow, the principal street is North High Street, backlands of the settlement along the North High with some settlement along Market Street, running Street. down to the Old Bridge and along the west bank of Later additions to the street plan of Musselburgh the Esk. The harbour stands apart to the west. These proper, such as Mansfield Gardens and Kilwinning two streets are separated by gardens or backlands. Terrace, appear to have been built on open gardens This basic layout is confirmed, with more detail, and backlands to the south of the town (illus 3). This on Roy’s Military Survey of Scotland (1747–55), area was largely undeveloped until the 20th century, which also shows a small strip of settlement on the with the exception of the Newbigging pottery and east side of the Esk to the north of Musselburgh’s the narrow strip of housing that lay either side of High Street, along the street that became Millhill. the road south to Inveresk. The High Street appears at the time of the survey to While the burgh has developed around its original have lain along the route of a minor spur road that street plan, the centre has seen some 20th-century rejoined the main road from the Old Bridge further development, particularly along the High Street to the east. The town layout, as seen in the 17th and west of Kilwinning Street, and in the plots to the 18th centuries, suggests strongly that Musselburgh rear of both frontages. The least affected area is proper did not develop directly beside the Old Bridge, the northern frontage east of Kilwinning Street, which appears to have been clear of buildings on while in Fisherrow, the North High Street between the east side of the Esk, but was probably always Lochend Road North and Bridge Street appears to focused on the area around the mercat cross and the retain much of its 18th- and 19th-century character High Street. (illus 2 & 3).

 4 SUMMARY OF KEY RESULTS

The type of work taking place in each area heavily tides, periodically added to the difficulties of the influenced the impact of the works on archaeologi- archaeologists! cal remains. In streets where the new mains were The final factor to be considered in assessing the installed using pipe-bursting or slip-lining tech- quality of the results is the human factor. The course niques very little archaeology was encountered, of the works was at times unpredictable or subject as these methods involved the excavation of small to short-notice changes of schedule in response trenches over the existing main. Likewise, the to conditions on the ground. Levels of experience replacement of house-to-main services required only among the 13 archaeological staff engaged in the small trenches over the old service, and the majority monitoring and recording also varied. Nevertheless of the deposits in these trenches were also disturbed. the project archive is coherent and consistent, and In contrast, archaeological remains were frequently provides a proper record of the nature, depth and encountered when the new mains were installed in extent of archaeological deposits beneath the road open-cut trenches, although in some cases the new surfaces in Musselburgh. pipe track lay directly alongside other services and either largely or completely in disturbed ground. The main open-cut trenches varied in depth between 4.1 Interventions around known sites 1m and 1.5m, and were usually around 0.5m wide. Other interventions depended on the depths of the Several interventions took place in the vicinity of existing services. known sites (illus 1, illus 5). The results of these There were several factors that affected the interventions are presented here, along with a brief quality of the data gathered. The most important, interpretation of their significance, and an assess- as mentioned above, was the lack of archaeologi- ment of the further potential of the area cal monitoring along much of the High Street. Archaeological deposits located in a few of the smaller house-to-main services in this area 4.1.1 The vicus indicated that the potential for locating significant remains in the open-cut trenches would have been Level with the northern boundary of the Scheduled high. A lesser factor affecting the data gathering area of the vicus (SAM 1182) trenching along the was, in some areas, the depth of the open-cut southern part of Inveresk Brae located a ditch trenches, which in places exceeded safe working approximately 6.5m wide and 1.5m deep (T200/012; depths and required recording to be done from illus 4). This feature contained a number of weath- the road surface. Along those parts of the High ering episodes (T200/007–010) along its southern Street that were monitored, variation in water edge, possibly indicative of the presence of a bank levels within the trenches, probably caused by the on the up-slope side. Fragments of glass, possibly

N S 15m OD 001

002 003

011 010 009 008 007

Cut 004

005

013 Cut 012

0 2.5 m

Illus 4 Section of Roman ditch on Inveresk Brae

 Roman, were recovered from the first weathering 4.1.2 The Newbigging pottery episode. Following a brief period of stability, in which fine clay (T200/013) accumulated in the base Open-cut trenching along the east side of Mansfield of the ditch, the feature was backfilled (T200/005). Avenue, towards the south end of the street, A fragment of Samian ware was recovered from uncovered a substantial 2.5m wide ditch or pit filled this deposit, which also contained fragments of with sand and large quantities of broken white china sandstone and lime mortar, possibly indicating (illus 5: T69/004). A large deposit of brick and mortar demolition in the vicinity. A shallow, sterile U- demolition rubble, again rich in white china, was shaped feature (T200/004) had been cut into the present in the northern end of the trench. Several final backfilling of the ditch, but yielded no dating other narrow-cut features crossing the northward evidence. continuation of this trench were probably furrows, The upper fills of the ditch were sealed by a dark the cultivation soil to which they were related brown buried soil (illus 5: T200/003). In the vicinity contained fragments of early post-medieval pottery. of the ditch, this deposit contained fragments of roof In 1987–88 a limited excavation at the site of tile and box-flue tiles, while further to the north it the Newbigging pottery located a sequence of kiln contained charcoal, oyster shell and small fragments bases, in addition to huge numbers of pottery sherds of bone. To the north of the ditch this deposit ran (Haggarty & McIntyre 1996). Unfortunately, the beneath a layer of redeposited sand and gravel, china discovered during the monitoring on Mansfield which while relatively thin in the area of the ditch Avenue, some 60–70 m to the west of the location increased to a depth of around 1m to the north as of the Newbigging pottery, was not sampled, as the the original ground slope fell away. potential significance of the deposit was not realised As mentioned above, Roman remains, including at the time. Without sherds from the watching hypocaust structures using box-flue tiles, were brief to compare with those known to have been encountered very near to this area during work at made at the pottery, it is impossible to say for sure Inveresk House in the 18th century. It is also inter- whether the deposits of china and rubble located on esting to note that during this work the owner of Mansfield Avenue bear any relation to activity at Inveresk House apparently filled in the east ditch the pottery. However, a hypothesis that could poten- of the fort, which had been used as the public road tially be confirmed by further work in this area to Inveresk from Musselburgh, and established is that the owners of the pottery may have made Inveresk Brae in its present location to the east of temporary use of the vacant High Street backlands this, in places cutting the road into the slope (de to the west of the pottery for the disposal of waste Cardonnel 1822, 164–5). The remains located in the pottery and rubble resulting from rebuilding at the course of the watching brief might be interpreted in factory. The Ordnance Survey town plans (1853 and relation to this information. The ditch seems to be 1894) give no indication of structures in this area, a fairly secure Roman feature, and possibly formed so it is possible that any activity either predated OS some sort of boundary or enclosure related to the mapping, or was very transitory in nature. fort or to the vicus. Other archaeological work in the properties to either side of Inveresk Brae has been limited in extent (Rogers 2002; Thomas 1988b) 4.1.3 The mill lade and this ditch appears to be a previously unknown feature of the vicus. Its location, prior to the levelling Trenching in Kerr’s Wynd cut across a 3.5m-wide effect of the made ground to the north, would have vaulted sandstone arch located directly beneath been at the top of a relatively steep slope. The buried the road surface (illus 5: T264/007). The arch rested soil overlying this feature, and extending down the on substantial sandstone and brick walls, and was hill, is probably the original post-medieval soil repaired in places with concrete and wooden sleepers. horizon prior to the establishment of Inveresk Brae. The interior of the arch was largely filled in with The Roman finds from this deposit and the shell, assorted demolition rubble and other detritus. A charcoal and bone may be material derived from large modern pipe was visible within this material, the 18th-century disturbance of Roman buildings in which on inspection by the contractors proved to be this area. The substantial deposits of made ground carrying a strong flow of water. to the north of T200/012 that seal the soil horizon A steady water supply was important to Mus- are probably related to efforts during construction selburgh, not only for the industries that produced of the Brae intended to reduce the severity of the goods for trade and export, but for the mills that slope. ground the corn for making bread. There is indirect The work on Inveresk Brae has demonstrated the evidence that the Esk was being harnessed for potential for Roman remains to survive beneath industrial purposes even before the end of the 13th road surfaces within the vicinity of the Scheduled century, and the town lade is historically attested areas. Further work at the head of the Brae, on from the 16th century when the existence of mills at Inveresk Village Road, was almost entirely located the foot of Kerr’s Wynd is also mentioned (Dennison over existing services. However, it is likely that & Coleman 1996, 19, 21). The existing mill lade undisturbed sections of the roadway in this area runs to the rear of the southern frontage of the High have similar potential to Inveresk Brae. Street, before crossing by the junction with Kil-

 Firth of Forth Fisherrow Links Beach Lane

Harbour 673000

T481/004 Downie place

Fishers Wynd Lochend Rd N

T264/007 T484/001 Linkfield Rd Millhill North High St River Esk Kerr’s Wynd 672800 Market St

Bridge St T128/006 T003/018 T128/008 Eskside West T010/027 Mercat

Market St Cross High Street KEY 672600

Location of medieval and later deposits T272/004

Mansfield Avenue Location of post-medieval and later deposits Location of Newbigging Pottery (OS 1st edition) Location of undated deposits T069/ 672400

Mill lade 004 Newbigging

Archaeological feature

Built-up area

N 672200

T200/012 0 250 m I nv

e

re

s

k

B

r a e 334400 334600 334800 333400 333600 333600 333800 333800 334000 334000 334200 334200 334400 334600 334800

672000

Illus 5 Location of key sites and archaeological deposits winning Street and running along the rear of the during an archaeological evaluation, although the burgage plots on the north side of the High Street. In sandstone arch was not breached (Mudie 2002). The the course of development within the burgh the lade town lade is quite clearly still a part of the modern has changed from a primarily open watercourse at infrastructure of the burgh, and it would probably the end of the 19th century (Ordnance Survey 1894) be unwise to compromise its structure should it be to one that is now largely culverted and piped. The encountered during any future archaeological work. structure located in Kerr’s Wynd, while difficult to However, as the flow of water now appears to be date in such a narrow slot, is part of a long tradition piped, it is possible that future monitoring exercises of lade bridges at this particular location. will provide the opportunity to obtain samples from A further section of the lade was identified on the base of the lade that could provide a date for the Kilwinning Place, to the south of the High Street, feature.

11 5 DISTRIBUTION OF ARCHAEOLOGICAL DEPOSITS

This section summarises the general archaeologi- stone-built box culverts. Deep plough-soils were the cal landscape within historic Musselburgh, and only deposits encountered along Carberry Road to Fisherrow (illus 5). The results have been tabulated the south of Inveresk. in order to give an indication for each area of the nature and date of the archaeological deposits located during the watching brief. 5.1 Musselburgh: High Street, Millhill, Kerr’s There were three main types of archaeologi- Wynd, Mansfield Avenue, Kilwinning Place, cal deposit located in the course of the watching Newbigging brief along the main streets of the burgh. The first is termed here ‘occupation deposit’ and generally Open-cut trenching was monitored on the High consisted of relatively thin dark brown, sandy Street between the mercat cross and East Port, sediment containing small fragments of shell, coal, and house-to-main services were recorded along charcoal and pottery, usually in low concentra- the length of the street. As mentioned above, tions. This type of deposit was mainly confined to archaeological monitoring of the project did not the main thoroughfares of the burgh and probably commence until substantial sections of open-cut formed during continuous use of these streets. It trench on the High Street had already been dug probably reflects an almost open-ended list of activi- and backfilled. ties that may have contributed to its makeup. Work The results of monitoring in this area are elsewhere has suggested that sediment deposition presented in Table 1 below. The only evidence for within medieval burghs is likely to derive partially medieval activity was located on the High Street. from the decay and mixing of midden deposits and In this area the early archaeological horizons imported building materials such as timber and turf. were generally sealed beneath redeposited beach Where these deposits exist in depth their presence gravels, possibly road surfaces, containing 18th- probably reflects intensive activity within the areas and 19th-century finds. Cut features were identified in which they are found (Carter 2001). Many of the in several trenches along the High Street (illus 5: occupation deposits located in the streets of Mus- T003/018, T010/027, T128/008). Although their selburgh are not deep, suggesting that the intensive original function is unknown, one of these features use of timber and turf constructions was focused (T010/027) was 0.5m deep and contained midden- away from the street frontage, with the added pos- type material and medieval pottery. These features sibility that the ‘occupation deposits’ within the seem unusual for a roadway, and may reflect earlier current roadways have been truncated. activity in these areas or changes in the street The term ‘midden-type’ is used here to refer to the layout. Similarly, a north–south-aligned mortar- second type of deposit, which was very rich in shells, bonded wall footing located in one trench on the many complete, and which also contained pottery. High Street (illus 5: T128/006) does not correspond This material often filled shallow cuts or appeared to any known mapped boundary and could reflect in discrete hollows or lenses. The sediment appeared alterations to the street frontage. largely unmixed and may represent deliberately A series of five east–west-aligned cuts 2m wide deposited material. and approximately 1m deep crossed Mansfield In the back-street areas of the core settlement Avenue between its north and south ends and (Mansfield Avenue, Kilwinning Place) very dark produced post-medieval pottery. These features soils containing occasional pottery fragments were were filled with the same cultivation soil horizon frequently encountered and have been termed that was prevalent throughout this area. The east– ‘cultivation soils’. The depth of these deposits was west-aligned mortar-bonded wall footing located on approximately half a metre, and appeared to show Mansfield Avenue illus( 5: T272/004) appears to cor- a single soil horizon overlying natural sediments. respond to the rear boundary of one of the burgage These deposits are likely to have formed at the plots shown on the OS first edition (illus 3). interface between the backlands of the burgh and the surrounding agricultural fields. Other cut features filled with homogenous 5.2 Fisherrow: North High Street, Market Street, sediments may have been boundaries or drainage Bridge Street features, and are summarised in the relevant section below. Open-cut trenching in this area exposed post- Further out from the burgh core, along the fringes medieval occupation horizons on North High Street, of Musselburgh, Fisherrow and in Inveresk, the Bridge Street, Eskside West and Market Street (Table only archaeological remains located were occasional 2). In general, the archaeological deposits consisted

12 Table 1 Distribution of archaeological deposits: Musselburgh

Midden-type Occupation Cut Cultivation Wall Culverts deposits features soil footings High Street M PM ND PM M ND ND Millhill ND ND Kerr’s Wynd ND ND Mansfield Av PM PM PM Kilwinning Place PM Newbigging ND ND

M Medieval PM Post-medieval ND No date

Table 2 Distribution of archaeological deposits: Fisherrow

Midden-type Occupation Cut Cultivation Wall Culverts deposits features soil footings North High St PM PM ND PM ND Downie Place ND Beach Lane ND Market Street ND Bridge Street PM PM ND Eskside West PM

M Medieval PM Post-medieval ND No date of a single horizon stratified between the natural deposits and cultivation soils were more common sands and gravels and modern road material, and than re-worked occupation deposits. This appears to included intermittent occupation deposits, patches fit well with this area’s status as backlands prior to of midden-type material and cut features crossing the 19th century. the trench. The cut features on North High Street Small trenches dug over the existing water main were filled with deposits similar to the homogenous on Market Street indicated the presence of undated deposits that underlay the modern road surface and occupation deposits in areas where settlement is their bases were rarely reached within the service depicted on the 17th-century maps. Neither the trenches. The presence of these features along a north–south-aligned mortar-bonded wall footing roadway is unusual. These may be later features located on North High Street (illus 5: T848/001) backfilled with redeposited occupation deposits, or nor one located on Beach Lane could be related to may represent activity pre-dating organised settle- mapped structures or boundaries. ment on the west side of the Esk. In contrast to the High Street in Musselburgh, On Bridge Street, which cuts across the former dated evidence for settlement in Fisherrow seems to backlands of North High Street, in situ midden be exclusively post-medieval and later.

13 6 ARTEFACTS AND THEIR DISTRIBUTION by Julie Franklin

6.1 Roman finds form hollow ducts to carry heat through the walls. It probably derives from the bath-house, as do the Several Roman finds came to light, though all were box-flue tiles from Cramond Roman Fort (Franklin fragmentary and probably redeposited. Not surpris- & Collard 2003). ingly, these were concentrated towards the south There are several unstratified sherds of tegula end of the area, in the vicinity of the Roman fort roof tiles from T200 and one unstratified lump of and settlement of Inveresk. They were found in brick (T199), none large enough to give informa- Trench 199 and Trench 200 towards the south end tion about size or form. The fabric varies from red of Inveresk Brae, at the north end of the Scheduled to light-red, with quartz and iron oxide inclusions, area, near the site of a possible bath-house. Neither sometimes with occasional mica and streaks of of these trenches produced any material of demon- white clay. These correspond to Crowley’s Inveresk strably later date. Fabrics 1 and 2 (Crowley forthcoming). The finds included one undecorated rim fragment of Samian ware from the backfill of the ditch (T200/005). This is from a plate or bowl possibly 6.2 Medieval and post-medieval finds of Drag. form 31, 37 or 38 or similar (see Oswald & Pryce 1920, Pl.XLV, Pl.LXXII, Pl.LXXV). These Medieval finds were limited to 11 sherds of White types were all produced in 2nd-century central Gaul. Gritty ware, produced in south-east Scotland Forms 31 and 37 are known in Scotland, being found between the 12th and 15th centuries. Sherds were at nearby Cramond (Dickinson 2003, 42). most common on the High Street (T30 and T39), Four fragments of clear glass were recovered though two redeposited fragments were also found from deposits associated with the weathering of across the Esk at Bridge Street (T659). the ditch cut (T200/007). These appear to be flat, Early post-medieval wares number 33 sherds, and thus are probably window glass. They are in and by contrast with earlier wares are concentrated very good condition, with only a slight laminating on the west side of the river on the North High effect on the surface of some. They could be post- Street. They include fragments of imported German medieval, but given the context, a Roman date is stoneware, delftware and slipware, from the Nether- possible. lands or England. Other finds are all of ceramic building material. Modern finds are naturally the most numerous. There is one sherd of box-flue tile (T200/003), with These include the usual types of utilitarian redwares, a knife-trimmed edge and combed, keyed surface. It creamware, bottle glass, clay pipes and pantiles. is of a hard-fired light-red gritty fabric, with a light- These are most common on Eskside West and North grey core. Box-flue tiles were stacked together to High Street.

(Approximate diameter)

0 10 cm

Illus 6 Sherd of Samian ware recovered from T200/012

14 7 discussion and Interpretation

The interventions around the three known sites have Musselburgh was quite clearly being bypassed to been discussed in the preceding section. This section the south by the main road running from the Old addresses the general archaeological landscape of Bridge by the 17th century. The bypassing may be medieval and later Musselburgh, as reflected by the original to the layout of the burgh, but could also results of the watching brief, and contrasts these be related to the post-medieval economic decline results with those from similar exercises in other of the burgh, when it was no longer worthwhile for Scottish burghs. travellers on the road to or from Edinburgh to pass through the town.

7.1 The archaeology of Musselburgh 7.1.3 Fisherrow 7.1.1 Distribution of deposits The date of settlement in Fisherrow has up until In situ archaeological deposits appear to survive now been unclear. It has been suggested that the well in the core of the historic burgh, and in harbour at Fisherrow began to replace a harbour- the vicinity of the vicus. Additionally, deposits age on the Esk in the medieval period, although probably related to the smaller satellite settle- historical sources only begin to refer to it by name ments such as Newbigging and the burgage plots in the 16th century. To date archaeological discov- on Market Street appear to survive, although eries have indicated only post-medieval activity in dating evidence was not recovered. Culverts and this area (ibid, 70). The evidence gathered by the other early drainage structures were generally monitoring in Fisherrow – where archaeological well preserved, and in places old property bound­ sampling was far more comprehensive than in Mus- aries may survive. selburgh – largely confirms the post-medieval date of significant settlement west of the Esk. Activity along the core North High Street dates only to the 7.1.2 Medieval Musselburgh 16th century or later, while the middens located on Bridge Street and Eskside West are of 17th- or The monitoring exercise provided a partial transect 18th-century date. The occupation deposits located through the core of medieval Musselburgh. The on Market Street – which correspond to the location fill of a cut feature to the east of the mercat cross of burgage plots on Adair’s map of 1682 – may be in this area (illus 5: T010/027) contained early of similar post-medieval date. It is possible that medieval pottery, and a layer of midden material these plots were only laid out west of the Esk after to the east of this contained both late medieval development along the High Street in Mussel- pottery and post-medieval roof tile. Combined burgh proper had run up against the river bank. with the cartographic evidence discussed above, The expansion of organised settlement west of the there is as much evidence for the location of the Esk probably reflects the growing importance of the burgh nucleus being along the current course of Fisherrow harbour to the townsfolk in this period. the High Street, as there is for a location in the The occurrence of imported pottery in this area vicinity of the Old Bridge. However, the occur- probably reflects a degree of international traffic rence of cut features in the areas monitored along into Fisherrow harbour, despite the dominance of the High Street suggests that the precise develop- nearby Leith. ment of this area remains to be fully understood. It is difficult to establish what the original focus for settlement would have been – the Old 7.2 Pipeline transects in historic burghs Bridge origin was suggested partly because of the potential for economic exploitation of the river The archaeological deposits located within the crossing (Dennison & Coleman 1996, 18) – but the core of Musselburgh are noticeably different in laying-out of formal burgage plots was probably a character from those found on similar monitoring reorganisation of existing settlement in an area exercises in Crail, North Berwick and Perth (Lowe with a long tradition of habitation. On its current 2001, Dingwall forthcoming, Glendinning 2002) par- alignment, the eastern end of the High Street is ticularly in the absence of stratified deposits. The far more suited for the positioning of burgage plots occupation deposits located within Musselburgh and than the west, as in this direction the available Fisherrow generally consisted of a single occupation land behind the northern street frontage becomes horizon of mixed date, with occasional more secure progressively shortened by the course of the Esk. pockets of material, such as shell midden filling pits.

15 In contrast, the monitoring of water mains renewal around more significant developments. However, in North Berwick located successive medieval and Perth is obviously a far larger town than Mussel- post-medieval surfaces, interspersed with deep burgh, and archaeological preservation within Perth layers of wind-blown sand, and rarely encountered town centre is excellent. underlying natural deposits. In Crail, monitoring The data from Musselburgh indicate that at either of the installation of a new sewage network again end of the High Street the archaeological deposits located in places successive medieval and later road were relatively intact and had been sealed by surfaces. redeposited beach gravels. Along North High Street, where modern development has generally been less intense, the archaeological horizon appeared in 7.2.1 Reasons for variation – population and places to have been truncated by the modern road development surface, but again the finds assemblage indicates that the surviving horizon is representative of the There are several probable factors lying behind the whole period between the 16th and 19th centuries. differences between the results of the three different monitoring exercises. The first, as has been stressed before, is methodological. It is entirely possible that 7.2.2 Reasons for variation – geography the most well-preserved and well-stratified medieval deposits within Musselburgh are located in that part It therefore seems likely that while the methodol- of the High Street west of the mercat cross, where ogy of the watching brief and modern development no open-cut trenching was monitored. However, the have had an impact on the survival of archaeologi- monitoring exercise was comprehensive throughout cal deposits in Musselburgh, additional factors are Fisherrow, where a deposit sequence from at least involved in generating the single mixed-date horizon as early as the 16th century could reasonably have that characterises the data from the watching brief. been expected, yet present was the same kind of The most striking difference with the results from single-horizon deposits as those encountered on the North Berwick is the lack of deep wind-blown sand. other bank of the Esk. It has been suggested that inundation by wind-blown Variation could also be attributed to the con- sand was a seasonal problem in North Berwick, and trasting histories of the three burghs. The most that many of the deposits seen beneath the roads obvious difference between Musselburgh, Crail represent attempts to reconsolidate the surface and and North Berwick is that of scale. Crail and, to a create metalled surfaces out of midden material lesser extent, North Berwick were both important (Dingwall forthcoming). Conditions in Perth, trading and economic centres in the medieval where deep stratified midden deposits also survive period, whilst the economy of Musselburgh seems beneath the road surface, could also be contrasted to have suffered through the burgh’s proximity to with Musselburgh – in this case anaerobic preserva- Edinburgh. However, over time this situation seems tion appears to have prevented the breakdown and to have reversed. The economic fortunes of the three reworking of archaeological deposits (Glendinning burghs diverged markedly following the medieval 2002, 96). period, with Musselburgh generally gaining pros- The geographic situation of Musselburgh is quite perity and becoming industrialised, while Crail and different from that of Perth and North Berwick. It North Berwick were reduced to relatively minor is further from the coast than North Berwick, and backwaters. In the 20th century Musselburgh has the seaward land is generally consolidated as links, far outstripped Crail and North Berwick in popula- rather than open beach as at North Berwick. This tion. The 2001 census gives the adult population in may be why sedimentation in the roadways seems to Musselburgh as 17,476, as opposed to 4,938 in North have been far less active, while unlike in Perth, the Berwick and 1,383 in Crail. The relative sizes in the waterlogging of archaeological deposits is probably three populations can obviously have correlations to far more periodic in Musselburgh. The combination the degree of modern disturbance of archaeological of soil conditions and a lack of active natural sedi- remains, particularly as the historic cores of these mentation in Musselburgh and the absence of built burghs remain their centres of social and economic road surfaces probably allowed for frequent distur- activity. In Musselburgh, this disturbance not only bance and mixing of material deposited in the streets, includes the construction of the modern shopping forming the homogenous sediments observed during buildings at the west end of the High Street, but also the monitoring. Processes such as street cleaning or extends to infrastructure development, such as the road resurfacing may also have contributed to the laying of services and improvements of the roadway. truncation of the resulting mixed-date archaeologi- Modern development in Musselburgh may have cal horizon, possibly accounting for its shallow depth truncated or obliterated archaeological horizons in most central areas of the burgh.

16 8 Conclusion

The recent work in Musselburgh has demonstrated As archaeological deposits are difficult to interpret the survival of archaeological deposits relating to in section in narrow slot-trenches, finds distribu- existing sites, Scheduled areas and the historic tions are probably the most valuable information burgh core. In general, extant horizons appear to that can be obtained from pipeline monitoring match closely with known areas of Roman and and the retrieval of securely stratified finds of all medieval settlement, but have been damaged periods must be the main priority for further work in places by the creation of modern roads and of this nature in Musselburgh, and in other historic services. Areas of particular sensitivity are the burghs where similar sedimentary sequences are High Street, North High Street, Newbigging and predicted. Inveresk Brae. A final point to make is that permitted develop- It cannot be denied that, had archaeological moni- ments such as water mains renewal can have large toring of the water mains renewal commenced sooner cumulative impacts upon archaeological remains. In in Musselburgh, the data would be far superior. As Musselburgh, the disturbance by existing services of the data from Musselburgh stand, generalisations this finite resource is already severe. Priority should given here on the development of the medieval burgh be given to ensuring that work within the roadways should be taken as models open to further testing. of historic burghs is rigorously controlled.

17 9 REFERENCES

Adair, J 1682 Map of Midlothian. Edinburgh. Holmes, N 2003 Excavation of Roman Sites at Bishop, M C 2002 ‘Inveresk Gate’ in Bishop (ed) Cramond, Edinburgh. Soc Antiq Scot Monogr 2002, 29–35. 23, Edinburgh. Bishop, M C (ed) 2002 Roman Inveresk: Past, Present Kimber, M 2004 Water Mains Renewal, Mussel- and Future. Arnatura Press, Duns. burgh, East Lothian: Data Structure Report of de Cardonnel, A 1822 ‘Description of certain Roman an Archaeological Watching Brief. Headland ruins discovered at Inveresk’, Archaeologia Archaeology Ltd unpublished client report. Scotica 2, 159–167. Knox, J 1816 Map of the Shire of Edinburgh. Carter, S 2001 ‘A reassessment of the origin of the St Edinburgh (surveyed 1812). Andrews “garden soil’’’, Tayside and Fife Archaeo­ Laurie, J 1763 A Plan of the County of Midlothian. logical Journal 7, 87–92. Edinburgh. Cook, M 2004 ‘How Mire: excavations across Leslie, A 2002 ‘The Roman fort at Inveresk’, in the cropmark complex at Inveresk, Mussel- Bishop (ed) 2002, 17–28. burgh, East Lothian’, Proc Soc Antiq Scot 134, Lowe, C 2001 ‘The distribution and survival of 131–60. medieval and later deposits in the medieval Crowley, N forthcoming ‘Building materials’, in burgh of Crail, Fife’, Tayside and Fife Archaeo- Leslie, A & Will, R ‘Excavations at Inveresk logical Journal 7, 93–109. Roman Fort, 1991–2001’, Proc Soc Antiq Scot. Lynch, M 1988 ‘The social and economic structure Dennison, E P & Coleman, R 1996 Historic Mus- of the larger towns, 1450–1600’, in Lynch, selburgh: The Archaeological Implications of Spearman & Stell (eds) 1988, 261–88. Development. Council for British Archaeology, Lynch, M; Spearman, M & Stell, G (eds) 1988 Edinburgh. The Scottish Medieval Town. John Donald, Dickinson, B 2003 ‘Samian ware’ in Holmes 2003. Edinburgh. Dingwall, K 2004 Water Mains Renewal, North Mair, C 1988 Mercat Cross and Tolbooth. John Berwick, East Lothian: Data Structure Report Donald, Edinburgh. of an Archaeological Watching Brief. Headland Mudie, G 2002 ‘Kilwinning Place, Musselburgh, East Archaeology Ltd unpublished client report. Lothian (Inveresk parish), mill lade’, Discovery Dingwall, K forthcoming ‘The archaeology of the and Excavation in Scotland 3, 38. streets of North Berwick and implications for the Old Statistical Account of Scotland 1791–99, volume development of the burgh’. SAIR, Edinburgh. 16 Inveresk, County of Edinburgh. Ewart, G & Triscott, J 1993 ‘Brunton’s Wireworks’, Ordnance Survey 1853 Musselburgh and its Discovery and Excavation in Scotland, 56. Environs. 1:1056 scale map (surveyed 1853). Franklin, J & Collard, M 2003 ‘Tile and brick’, in Ordnance Survey 1894 Musselburgh. 1:500 scale Holmes 2003. map (surveyed 1893). Glendinning, B 2002 ‘Tayside Mains Replacement, Oswald, F & Davies Pryce, T 1920 An Introduction to Perth’, Discovery and Excavation in Scotland, the study of Terra Sigillata. Longman, London. 96–97. Rogers, I 2002 ‘Eskgrove, Inveresk: a watching brief Haggarty, G & McIntyre, A 1996 ‘Excavation and in the vicus’, Bishop (ed) 2002, 36–40. watching brief at Newbigging Pottery Mussel- Roy, W 1747–55 Military Survey of Scotland. burgh, East Lothian’, Proc Soc Antiq Scot 126, Thomas, G 1988a ‘Excavations at the Roman civil 943–62. settlement in Inveresk, 1976–77’, Proc Soc Antiq Hastie, M 2001 ‘118 High Street, Musselburgh, Scot 118, 139–76. East Lothian (Inveresk parish), watching brief’, Thomas, G 1988b ‘The re-excavation of the Inveresk Discovery and Excavation in Scotland, 33. Hypocaust’, Proc Soc Antiq Scot 118, 177–79.

18